Desempenho em leitura e resolução de problemas matemáticos na Prova Brasil

• Autor(a) principal: Matos, Andrea Maria dos Santos
• Data de Publicação: 2018
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Repositório Institucional da UFS
• Texto Completo: http://ri.ufs.br/jspui/handle/riufs/9199
• topic: Educação; Educação básica; Compreensão na leitura; Ensino de matemática; Avaliação educacional; Avaliação de estudantes; Prova Brasil; Resolução de problemas; Basic education; Reading comprehension; Brazilian assessment policy; Problem solving; CIENCIAS HUMANAS::EDUCACAO

Brazilian assessment policy-Prova Brasil is a large – basic - scale evaluation of basic education applied to the 5th and 9th years of elementary education in Brazilian public schools, which aims is to measure performance in Mathematics and Portuguese Language, in problem solving and reading skills. To solve mathematical problems, one must read the statement and understand what is required to develop mathematical procedures. Therefore, there is a relation between the performance in the Brazilian Course of Portuguese Language and Mathematics, which is the objective of this research study of this thesis, developed in three stages. The first step was to identify the cognitive boundaries between reading and mathematics. From the pedagogical point of view, the two disciplines are sealed and separated. From the cognitive point of view, although the two areas of knowledge access distinct neural circuits, both share processes of representation and retrieval of symbolic information, attention, functional memory and cognitive control, which shows the approximation in terms of cognitive processing. The second stage consisted of the elaboration of a standardized test according to the Brazilian Course, in order to allow the monitoring of individual student performance. The tests were elaborated from the available items on the website of the Pedagogical Devolutive Platform of Inep; we chose two selection criteria using the proficiency, scoring levels and the Reference Matrix topics of each area / series evaluated; we distribute the items according to the stratification of difficult levels as: easy, medium and difficult; we tested the items to identify the degree of difficulty in three parameters: discrimination, probability of randomization and degree of difficulty, following the Precepts of Item Response Theory. After the validation of the elaborated test, we moved to the third stage, which objective was to compare students' performance in reading and problem solving. There was a total of 227 students who participated of this research from the 6th and 9th years of an elementary school in two public schools in Sergipe (state school and federal school), students’ average aged between 10 to 19 years old. The data collection procedures was evaluated by the research ethics committee, CAAE 2,008,797. The data was submitted to statistical analysis. Reading performance is associated with problem solving performance, presenting a strong and significant correlation for the 6th year (r = 0.62 and p = <0.001) and for the 9th year (r = 051 and p <0.001). The results indicate that the highest proficiencies in Mathematics correspond to the higher levels in Portuguese Language, for the 6th and 9th years. In the comparison between the performance in Mathematics and the Portuguese Language questions that required basic levels of proficiency, that is, those presented more elementary and more direct statements, there is no significant difference statistically. However, for more complex levels items, which skills are required to establish relationships between the textual elements and the problem statement, we observe the influence of the level of the proficiency Portuguese language in problem solving performance. We conclude that there is a relationship between performance in the Brazilian Mathematics and Portuguese Language Test, more specifically, in the relation between problem solving and reading proficiency, which points to the necessity of the articulated pedagogical work between the subjects for the effectiveness in the results.